Monday, December 26, 2016

Analyzing KHCB's Financials -- The Weakest Link: Profit Rate Risk

Paging Anne Robinson
Continuing our analysis of KHCB to examine its potential to support GFHFG's new strategy ..

Quick Summary (for the Twitter Generation): Profit Rate Risk (AKA Interest Rate Risk) appears very high at KHCB.  Thus, a poor "fit" with GFH's new strategy.

Let's start by taking a look at yield on KHCB's loan portfolio yield to set the stage.

Portfolio Yield

Sometimes banks don’t book “air” for AIR (Accrued Interest or "Income" Receivable).  Or if they realize or are forced to realize they have, they reverse it out.  When they do this, they generally don’t provide details.  But, if the amounts are large enough, the gross yield on the lending portfolio will decrease significantly.  Lower rates on renegotiated loans and non-accrual on income in part or full on some loans can also depress gross yield.  Trends in portfolio yield can also reveal if a bank is taking more risk relative to its peers.   

Gross yield—which is the amount of interest owed by borrowers—is best because trends are not masked by changes in cost of funding (COF) as they are with net yield.   However, one does need to keep an eye on COF.

Why?

Because bank loans are generally priced over defined COF typically against bank benchmarks, e.g., LIBOR, as a proxy for the lending bank’s cost of funds.   Note actual funding of loans is independent of their price setting mechanism.  There is nothing to prevent a bank from pricing a loan over three month Libor and using overnight funding or more than three month funding.  That mismatch is how a bank’s Treasury earns its keep.

If we’re using gross yield (margin + contractual cost of funds) to identify whether changes are due to credit problems or other causes, then we need to be able to isolate the impact of COF to determine if COF is driving changes in gross yield through the loan pricing mechanism.  

The first step is determining COF on bank placements (deposits). If financials provide a breakdown of COF for bank placements separate from customer placements, the analyst’s job is easier. When they do not, analysis is a bit more difficult as the reported COF includes both bank and non-bank deposits.  KHCB only provides a total COF which includes both bank and client placements. In KHCB’s case, analysis is complicated further because the relative percentages of bank and client deposits changed over the period analyzed.

The second step is determining the contractual pricing basis of assets.  This is usually in the “risk management note” near the end of the financial statement.  The “Islamic” equivalent of a conventional bank’s “interest rate risk” is “profit rate risk.”  That is, is the bank making loans that reprice every three months?  Every 12 months?   Every 240 months?

Some technical notes for the charts below:
1.   Portfolio yield is on “Financing Assets” and “Assets for Leasing” because gross revenue is given as a single figure for both.  Again not ideal from the analyst’s perspective. 
2.   COF is based on bank placements and placements from clients (non-bank institutions and individuals) because there was no breakdown of the components. 
3.   Current accounts are excluded because they are not relevant to loan pricing as explained above. 
4.   Percentages are calculated using a simple average (beginning of period and end of period) for both the loan portfolio and the “deposits”. 
5.   Contractual repricing data is as of FYE (31 December) for the years shown as per respective ARs. For 2015 see AR 2015 Note 34 page 75.
KHCB Lending Yield and COF Analysis

3Q16
2Q16
1Q16
FY15
FY14
FY13
FY12
FA & LA
5.80%
6.34%
6.30%
5.91%
6.84%
6.63%
7.57%
Depo COF
2.37%
2.26%
2.18%
2.51%
2.81%
3.24%
2.65%


Contractual Repricing of FA & LA Portfolios -- KHCB and BIsB
KHCB
Months
Up to 3
3 to 6
6 to 12
12 to 36
Over 36
2015
12%
4%
9%
20%
54%
2014
16%
7%
6%
20%
52%
BIsB
Months
Up to 3
3 to 6
6 to 12
12 to 36
Over 36
2015
6%
2%
5%
15%
72%
2014
14%
1%
6%
14%
64%


Contractual Repricing of FA & LA Portfolios – QIB
Months
Up to 3
3 to 12
12 to 60
60+
2015
41%
28%
22%
9%
2014
47%
27%
24%
3%

Comments on Portfolio Yield

1.   As the second chart above shows, KHCB and BIsB are largely fixed rate lenders.  So it’s no surprise that portfolio yield is not correlated with COF.
2.   Qatar Islamic Bank was “summoned” to this discussion to show what appears to be a more commercial bank lending portfolio. That is, heavily weighted to one year and under.  QIB owns to an effective profit rate of 4.2% (2015) and 4.32% (2014).  Not bad.  A natural question is what risks KHCB is running for an extra 1.6% and whether the risk/reward tradeoff is prudent. 
3.   Given this, KHCB and BIsB’s yield changes are driven by credit “events” on existing loans and leases (nonaccrual, reversal of AIR) and to a lesser extent by changes in market pricing for new loans and leases.  Because both banks are making long dated loans, the portfolio turns over slowly.  New financings are relatively small compared to the existing “book”. Thus, new financings only marginally affect gross yield.  If the portfolio “suddenly” increased or decreased (write-offs for example) by a large amount, this could affect net portfolio yield if current market rates were substantially different than those on the existing portfolio.  
4.   From FYE 2015 to 3Q16, KHCB’s FA & LA portfolio increased BHD 64 million point to point. Looking at simple average outstandings by quarter in 2016, the increase in 3Q16 was roughly BHD 26 million.  As the below table shows that new business would have to have been booked at less than a zero yield to cause a 50 bp drop in overall portfolio yield. That seems unlikely. 
5.   Thus, the drop in yield in 3Q16 is almost certainly credit related, e.g., non-accrual perhaps accompanied by reversal of previous period accruals, renegotiation of loans as a lower rate, or a combination of these factors. 
6.   If KHCB had maintained the 6.30% yield of 1Q16 and 2Q16 in 3Q16, gross interest revenue would have been roughly BHD0.560 million higher.   
7.   It doesn’t seem likely that the 3Q16 BHD2.2 million specific provision is solely responsible for the drop as nine months’ interest would be roughly BDH0.100 million using an average 6.3% portfolio yield.    
8.   So what would it take? 
9.   Assuming the yield should be 6.3%, the table below summarizes the single period amounts and rates at which they would have to be accrued (0% to 5%) to reduce the overall portfolio yield to 5.80%.  Note this analysis is for the full 50 bps drop.  Thus, it’s an analytical shortcut, ignoring Q1 and Q2 potential reversals on the 3Q BHD 2.2 provision and perhaps further back into 2015 if there was related AIR.  While a shortcut, it’s close enough for directional analysis.  Also note the required change is calculated using a simple average portfolio outstandings, e.g., (2Q+3Q)/2.   
10. In guesstimating probabilities of the above scenarios, AA thinks that zero accrual scenario is more likely that a massive renegotiation of loans to 5%.  Or partial accruals at the intermediate rates.   

5.80% Portfolio Yield "Required" Accrual Scenarios
Accrual Rate
0%
1%
2%
3%
4%
5%
BHD Millions
32
38
46
60
87
153
% FA & LA
7%
8%
10%
14%
19%
34%

 Comments on Profit Rate Risk    

1.   Profit Rate Risk (PRR) is the “Islamic” equivalent of conventional bank Interest Rate Risk a measure of the exposure of a bank to movements in interest rates.  If assets and liabilities are matched in terms of currency, actually repriced according to contractual repricing, and maturities then the bank has no interest or profit rate risk. It has locked in its net interest margin for the life of the loan absent credit events
2.   Banks generally don’t match 100% because they can make additional revenue through mismatching, typically borrowing short and lending long, assuming a “normal” upward sloping yield curve. 
3.   Another key factor and one which AA suspects is at play here is that typically sources of matching long dated funding are limited or non-existent.  Banks might want to match fund but they can’t.  Banks have three choices.  Make the fixed rate loan and short fund.  Refrain from making fixed rate loans for these tenors. Or make floating rate loans that reprice frequently.   An example would be a ten year loan that is priced off six-month Libor. 
PRR (IRR) comes in two flavors:  income risk and asset risk.  

1.   Income Risk:  If a bank matches as described above, and locks in its margin, changes in interest rates (COF) do not affect its income.  If it does not, it runs the risk that funding costs will exceed the earnings on the asset.  Think of First Penn or UBAF Arab American (NYC).  Both purchased long dated “riskless” US Treasury securities but funded with short dated money in the repo market.  As rates rose, first profit evaporated and then earnings went negative.  Both foundered. 
2.   Asset Price Risk:  With a fixed rate instrument, if market rates go up, the value of the instrument goes down.  Duration and convexity are measures of this risk--most familiar to bond investors.  Because loans and leases are in KHCB’s “banking book”, they are carried at historic cost less impairment for credit reasons.  IFRS doesn’t require that banking book assets be fair valued with losses or gains through income or directly in equity.   
3.   Risk Management:  In discussing management of market risks in its 2015 AR (page 73), KHCB notes that there are Shari’a compliant foreign exchange risk management transactions available but says nothing about the availability of profit rate risk management transactions.  That suggests as a practical matter there are none.  Further on page 75, it states “Overall non-trading profit rate risk positions are managed by Treasury department, which uses short term investment securities, placement with banks and placement from banks to manage the overall position arising from the Bank’s non-trading activities.”   Frankly, AA doesn’t see how KHCB could be using short term instruments to effectively hedge against the long term interest rate risk it appears to be carrying—up to 20 years!  Perhaps, a treasurer or trader reading this could set AA “straight”. 
4.   So far we’ve looked at half the PRR picture: yield on assets.  KHCB’s funding structure is the second and critical piece of PRR.   Below is a chart developed from KHCB’s 2015 AR Risk Management Note 34 page 75.
KHCB Funding Gap Analysis - FA & LA Portfolio
Months
Up to 3
3 to 6
6 to 12
12 to 36
Over 36
2015
FA & LA
47
17
34
78
211
Funding
96
45
70
43
112
GAP
49
28
36
(35)
(99)
% Cover
204%
267%
205%
55%
53%
2014
FA & LA
53
23
21
66
176
Funding
72
55
68
127
1
GAP
19
32
47
60
(175)
% Cover
136%
236%
325%
191%
1%

Technical Notes: 
  1. I’ve deducted both the related assets and funding for bank placements taken and investment in sukuk shown in Note 34 page 75.  The only other profit rate sensitive assets are FA & LA.   Thus, we have a “clean” picture of how the lending portfolio is funded. 
  2. Note that a negative “Gap” number means that there is insufficient funding for the respective time “bucket” of assets.  Such “gaps” are a measure PRR. 
  3. The longer the life of a fixed rate asset the longer the potential period of negative funding costs.  Therefore, negative gaps in longer dated maturities have more PRR than in the shorter maturities as a general rule.  
  4. Also while IFRS doesn’t require that these “banking book” assets be revalued for income or balance sheet purposes, their fair value must be reported in a footnote.  If the damage is severe enough, depositors and creditors may take “flight”.   
Comments on Funding Gap and Profit Rate Risk Analysis:   

1.   At first glance KHCB’s 2015 gap position is an improvement from 2014.  One might argue that the BDH 111 million in total equity covers the gap in the final maturity “bucket” shown.   
2.   The analytical problem is that we have no way of knowing the tenors of deposit funding versus assets.  The “over three years” category covers a wide range of possibilities.  Funding may be for 37 or 120 months. Assets may be for 240 months or 36 months and three days.  Note 34 doesn’t contain enough information to determine what the gap is.  Another criticism and suggestion for changes to reporting standards.  When an institution engages in long dated transactions, it should provide more detailed information on longer tenors. 
3.   IFRS may not have provided the answer.  However, we can thank the good folks at the Basel Committee on Banking Supervision for their Pillar III requirements. 
4.   AR 2015 Note 3.4.6 on page 88 provides residual maturities, not profit rate risk gap maturities.  However, we can use this information to form an opinion about the maximum tenors that KHCB extends on financings and the length time that KHCB is exposed to profit rate risk.  The former should be shown by the maximum tenor in the note.  The latter by the residual maturities themselves as portfolio level residual maturities will tell us how long it will take for outstandings to run off. Note with a fixed rate lender like KHCB, repricing does not take place or is infrequent.  Again a reporting deficiency by use of the “over three years” category. 
5.   Sadly, Basel doesn’t require a similar note on funding, but we can look at the funding gap “over three years category” and form an opinion on likely tenors of deposits based on the “market”. And as discussed below, we can look at additional notes to confirm or deny that opinion. 

Basel III Residual Maturity Note  -- FA & LA

Up to 1Y
1 to 3Y
3 to 5Y
5 to 10Y
10 to 20Y
20Y+
BHD Millions
99
78
66
90
44
11
Percent
25%
20%
17%
23%
11%
3%


1.   At the beginning of Note 3.4.6, KHCB notes that maximum exposures are 7 years for corporate and 25 years for retail borrowers, unless the Board approves exceptions.   Note that leasing accounts for all the exposure over 20 years. 
2.   Looking at the chart immediately above it seems unlikely that KHCB has found matching funding for such long tenors.  Who is placing 5 year, much less 10 or 20 year money with Islamic banks or other banks?  What depositors are placing 10 or 20 year money with regional commercial banks? How many depositors are placing that sort of money with investment grade global banks? 
3.   But there’s another way to get an insight into deposit tenors.  Note 8.3 on page 98 provides a breakdown of “interest” paid on IAH for 2015 and 2014.  The absolute amount of interest paid falls off dramatically after 12 months. Minimal amounts of “interest” are shown for 18 and 24 month Mudharab accounts in both 2015 and 2014.  How minimal? Less than 1% of the total interest paid to IAH holders. That suggests deposits in the 12 to 36 month “bucket” are largely in the 12 month category. And thus despite the apparent “match” there is a funding gap in this “bucket”.
4.   But where are the 36 month and over IAH deposits shown in Note 34 as representing 99% of the PRR Gap match funding in this bucket.  This is the most risky time period for PRR. Strangely there is no category for these deposits.  Since this maturity “bucket” holds over BHD 112 million or over 30% of all IAH deposits, it seems strange that there isn’t a separate category.  It’s certainly “material” by typical accounting standards. 
5.   There is a category for VIP Mudharab accounts.  Could this be the over 36 month category? Doesn’t look like it. On the following page these accounts are being paid 2.00% which is lower than the 6 month Mudharab “interest” of 2.69%.   Doesn’t seem like a preferential rate for VIPs but then the tenor is unknown because it is unspecified.  Maybe a variant of qard hassan?  AA doesn’t think so.  Also the total interest paid to the VIP IAH is 14% of total interest paid.  It would seem that interest for the more than 36 months deposits should be at least 30% of the total paid in line with their percentage of total IAH deposits.  But it is not. 
6.   AA has no explanation for either #4 or #5 above, and thus the tentative conclusion is that KHCB’s profit (interest) rate risk gap remains substantial and there is significant income risk
7.   In an environment where rates are likely to increase driven by the US FRB, KHCB is likely to face earnings pressure (another euphemism) as rates rise.  Not exactly a good strategic "fit" for GFH Financial Group which is looking for stable income not heavily influenced by global factors.

No comments: